Greta de Groat wrote:
I do think that the ideal situation is one where one has both
human-readable data and identfiers for many of the data elements. That
way you have the best of both worlds. Look at CCO, many of the
instructions also follow this path. I think it is a good thing for the
rules to say that this is a desirable state when possible.
But remember that this is not always going to be possible. For one
thing it will take more time to input this information--and how many
times over the years have you heard griping about having to input the
"redundant" information in fixed fields (not that this information was
ideally coded, but the impulse for machine processing was there).
Administrators will not see this as a simplification.
Greta, some time could be saved by embracing identifiers. Rather than
seeing identifiers as something you input, you could see them as sources
of information. For example, there is no reason for anyone to have to
key a publisher name for a modern book with an ISBN -- the identity of
the publisher is inherent in the ISBN and that information could be
system supplied (using a barcode scanner). But this would mean giving up
the idea that the publisher name must be transcribed from the piece. It
is that rule that is forcing us to continue to key the publisher name
rather than get that information from a readily available source.
Many people were quite affronted when the Future of Bibliographic
Control report suggested getting as much data as possible from the
publishers. There is data that they obviously have and presumably could
pass on. And a simple identifier like the ISBN is actually a window to a
whole host of information that could be used to populate a bibliographic
record.
kc
--
-----------------------------------
Karen Coyle / Digital Library Consultant
[EMAIL PROTECTED] http://www.kcoyle.net
ph.: 510-540-7596 skype: kcoylenet
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------
